Excessive release of host inflammatory mediators contributes directly to the pathogenesis of sepsis and septic shock. Mitogen-activated protein kinases (MAPKs) are serine/threonine intracellular signaling proteins that have a central role in this inflammatory response. These MAPKs are activated by phosphorylation and mediate gene expression and other cellular functions in response to extracellular signals. Three major MAPKs in mammalian cells are p38, extracellular signal-regulated kinase (ERK), and c-jun N-terminal kinase (JNK). Each of these MAPKs regulates inflammatory mediator production in response to lipopolysaccharide (LPS), a bacterial toxin closely associated with the pathogenesis of gram-negative bacterial pneumonia and sepsis. These mediators include cytokines, chemokines, nitric oxide, reactive oxygen species, and prostaglandin metabolites. Based on substantial data supporting a pivotal role for MAPKs in inflammation, there is growing interest in the therapeutic application of selective inhibitors of these proteins. However, the application of such agents in sepsis has been primarily tested in in vivo models employing LPS challenge only. Since many of the mediators MAPKs regulate participate in protective innate immune or host defense responses as well as inflammatory injury, their inhibition could be harmful during bacterial infection. Data from a study we conducted with Bacillus anthracis lethal toxin (LeTx) supports the possibility that there may be risks as well as benefits with MAPK inhibition. The lethal factor (LF) component of LeTx is a zinc protease that cleaves and inactivates several different members of the mitogen-activated protein kinase (MAPKKs) family and their downstream targets including p38. We showed that when LeTx was administered in sublethal doses to animals challenged with lethal LPS, inflammatory cytokine and nitric oxide release were reduced and hemodynamic function and survival improved. However, similar pretreatment in animals challenged with intratracheal E. coli, while also reducing cytokine and nitric oxide release, decreased survival. These findings suggested that nonselective inhibition of these MAPKs may be harmful with live bacteria. Selective inhibition of these MAPKs may still be beneficial however.[unreadable] [unreadable] SB203580 4-(4-flurophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)imidazole is a pyridinyl imidazole that has been employed extensively as a selective inhibitor of p38, in both in vitro and in vivo models of inflammation. This agent competes for the ATP binding site on p38 and inhibits its phosphorylation and activation. In in vivo models, administration of this agent via several different routes, as well as both before and after inflammatory stimulus, has been shown to inhibit cytokine and nitric oxide release, leukocyte trafficking, and organ injury. This agent and a related one have also been shown to improve survival in LPS challenged models. Although the results of such studies have provided a major rational for the use of p38 inhibitors during states of inflammation such as sepsis, an extensive literature search has revealed only one published report (12) assessing the effects of this or related agents in an animal model of sepsis employing live bacterial challenge. This was a very limited study in mice undergoing cecal ligation and puncture (CLP) and did not incorporate antibiotic support or microbiological assessment. The results of this study did suggest however that the effectiveness of SB203580 was increased when its administration was delayed for 12 h after the onset of sepsis. [unreadable] [unreadable] We previously developed a mouse model of E. coli pneumonia in which antibiotics and fluids were shown to synergistically improve survival as they are believed to do clinically. This model is now being used to test the effects of SB203580. In initial experiments we have found that doses of this agent previously reported to be beneficial in very limited studies, actually appeard harmful when administered either 1 h before or 1 h after bacterial challenge in the mouse model. Delaying treatment for 12 h is did not result in harmful effects but was also not beneficial. Reducing the dose of SB203580 3 logs doses resulted in beneficial effects with the agent, but only if it is given prophylactically. Later treatment had no benefit. In subsequent studies we have now found that while high and low prophylactic doses of SB203580 have similar effects on myocardial function as assessed by echocardiography, their effects on inflammatory cytokine release differed. Importantly, while low doses were associated with increased cytokine release early and decreases later, high doses produce uniform reductions. Thus, the harmful effects of high but not low doses of SB203580 may relate to differential expression of host mediators that are important not only for inflammatory tissue injury but also for host defense as well. Additional studies are ongoing investigating the effects of SB203580 on P38 levels.